محمد جمعه

Python automation scripts are commonly employed to streamline routine operational tasks, including file han‎dling, data preparation, reporting, an‎d lightweight system integration. These scripts are often developed under time constraints an‎d evolve in an ad hoc manner, leading to the accumulation of technical debt, such as inconsistent formatting, lint violations, high cyclomatic complexity, an‎d monolithic functions. These factors diminish maintainability an‎d elevate both the cost an‎d risk of subsequent changes. This thesis introduces a structured, task-specific framework for refactoring Python automation scripts through a metrics-guided pipeline that links baseline measurement, prioritization, tool-assisted transformation, stan‎dardized formatting, an‎d post-refactoring verification into a repeatable workflow. The framework incorporates established static analysis an‎d refactoring tools, including code formatters, linters, an‎d complexity analyzers. It generates machine-readable artifacts that document the configuration, detected issues, applied transformations, an‎d pre- an‎d post-quality deltas, ensuring reproducibility an‎d auditability. A prioritization policy is presented to direct refactoring efforts toward scripts with the highest maintainability risk by combining complexity indicators an‎d static diagnostics. The proposed approach is eva‎luated on a corpus of automation scripts, with pre- an‎d post-refactoring results compared using lint an‎d complexity metrics. The findings demonstrate substantial improvements in code quality, including the elimination of lint violations at the corpus level an‎d significant reductions in cyclomatic complexity an‎d maintainability risk in high-priority scripts, while maintaining verification gates to detect regressions. In summary, this thesis shows that automation-script refactoring can be transformed from an informal cleanup activity into an evidence-driven, repeatable engineering process suitable for operational environments.

اسكريپت هاي اتوماسيون پايتون , ريفكتورينگ خودكار , ريفكتورينگ مبتني بر معيارها , كيفيت نرم افزار , پايپ لاين ريفكتورينگ

Python automation scripts , automated refactoring , metrics-guided refactoring , software quality , refactoring pipeline

Mohammed Taha Jumaah

Dr. Mehrdad Ashtiani